Aero-engine blades crack detection is one of the important tasks in daily ground maintenance, crack is a kind of texture feature, due to the random distribution, irregular shape and vague characteristics, which is still a challenging task to realize automatic detection in working environment. A detection model based on the Yolov4-tiny is proposed that is universal and focuses more on the characteristics of cracks, and it is implemented in embedded device. First, in order to distinguish the cracks and noises, an improved attention module is introduced into the backbone of Yolov4-tiny to enhance the model's capability to focus on crack areas; second, in order to improve the effect of multi-scale feature fusion, the bicubic interpolation is implemented in upsampling module; finally, in order to solve the redundant detection results of bounding-boxes in crack areas, the optimized non-maximum suppression method is proposed to make the detection results better corresponding to the groundTruth. The robustness of proposed detection model was demonstrated by evaluating varying lighting and noise images. The average precision on integrated datasets is 81.6%, which outperforms the original Yolov4-tiny by an increase of 12.3%.
The “low, slow, and small” target (LSST) poses a significant threat to the military ground unit. It is hard to defend against due to its invisibility to numerous detecting devices. With the onboard deep learning-based object detection methods, the intelligent LSST (ILSST) can find and detect the ground unit autonomously in a denied environment. This paper proposes an adversarial patch-based defending method to blind the ILSST by attacking its onboard object detection network. First, an adversarial influence score was established to indicate the influence of the adversarial noise on the objects. Then, based on this score, we used the least squares algorithm and Bisectional search methods to search the patch’s optimal coordinates and size. Using the optimal coordinates and size, an adaptive patch-generating network was constructed to automatically generate patches on ground units and hide the ground units from the deep learning-based object detection network. To evaluate the efficiency of our algorithm, a new LSST view dataset was collected, and extensive attacking experiments are carried out on this dataset. The results demonstrate that our algorithm can effectively attack the object detection networks, is better than state-of-the-art adversarial patch-generating algorithms in hiding the ground units from the object detection networks, and has high transferability among the object detection networks.
Infrared imaging is widely used due to its penetration capability to operate under many weather or lighting condition. However, due to the far distance of aerial view, feature blur, and the scarcity of aerial infrared data, the detection of small infrared targets on the water surface remains a challenging problem. In response to the problem of unclear features, we propose the spatial feature weighting method based on 2D Gaussian distribution. This method increases the weight of the target area by adaptively adjusting the feature activation. Secondly, for the problem of rare aerial perspective infrared data, we propose the cross‐spectral data migration method. By introducing the domain difference loss function to optimize the pseudo‐label selection process, the range of target domain distribution is expanded, and the adaptability of the detector is improved. Finally, in response to the problem of underfitting caused by category imbalance in transfer learning, we propose the class balancing method that effectively reduces the false detection. Extensive experiments were conducted on both benchmark datasets and the self‐built dataset to evaluate the effectiveness and robustness of our method. The proposed method was evaluated with different models and various scenarios, and the results demonstrated the effectiveness.
Template matching is the fundamental task in remote sensing image processing of air- and space-based platforms. Due to the heterogeneous image sources, different scales and different viewpoints, the realization of a general end-to-end matching model is still a challenging task. Considering the abovementioned problems, we propose a cross-view remote sensing image matching method. Firstly, a spatial attention map was proposed to solve the problem of the domain gap. It is produced by two-dimensional Gaussian distribution and eliminates the distance between the distributed heterogeneous features. Secondly, in order to perform matching at different flight altitudes, a multi-scale matching method was proposed to perform matching on three down-sampling scales in turn and confirm the optimal result. Thirdly, to improve the adaptability of the viewpoint changes, a pixel-wise consensus method based on a correlation layer was applied. Finally, we trained the proposed model based on weakly supervised learning, which does not require extensive annotation but only labels one pair of feature points of the template image and search image. The robustness and effectiveness of the proposed methods were demonstrated by evaluation on various datasets. Our method accommodates three types of template matching with different viewpoints, including SAR to RGB, infrared to RGB, and RGB to RGB.
In this paper, a visual navigation method based on binocular vision and a deep learning approach is proposed to solve the navigation problem of the unmanned aerial vehicle autonomous aerial refueling docking process. First, to meet the requirements of high accuracy and high frame rate in aerial refueling tasks, this paper proposes a single-stage lightweight drogue detection model, which greatly increases the inference speed of binocular images by introducing image alignment and depth-separable convolution and improves the feature extraction capability and scale adaptation performance of the model by using an efficient attention mechanism (ECA) and adaptive spatial feature fusion method (ASFF). Second, this paper proposes a novel method for estimating the pose of the drogue by spatial geometric modeling using optical markers, and further improves the accuracy and robustness of the algorithm by using visual reprojection. Moreover, this paper constructs a visual navigation vision simulation and semi-physical simulation experiments for the autonomous aerial refueling task, and the experimental results show the following: (1) the proposed drogue detection model has high accuracy and real-time performance, with a mean average precision (mAP) of 98.23% and a detection speed of 41.11 FPS in the embedded module; (2) the position estimation error of the proposed visual navigation algorithm is less than ±0.1 m, and the attitude estimation error of the pitch and yaw angle is less than ±0.5°; and (3) through comparison experiments with the existing advanced methods, the positioning accuracy of this method is improved by 1.18% compared with the current advanced methods.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.